The Journal of Specialised Translation Issue 30 – July 2018

A typology of translation solutions

Anthony Pym, Universitat Rovira i Virgili / University of Melbourne

ABSTRACT

An eight-term pedagogical typology of translation solutions has been compiled and taught in
two Masters classes, one in the United States and the other in South Africa. The results
suggest that the typology is robust enough to be pedagogically effective in the two
situations if and when the teaching stresses a series of points: 1) the nature of its “problem-
solving” premises has to be explained carefully, 2) the typology should be presented as a list
of ways to address problems that cannot be solved using the norms of standard languages
or “cruise” mode translation procedures, 3) it should be presented as being open-ended,
inviting new solutions and new combinations of the main solution types, 4) its theorisation
should be kept as simple as possible, in the interests of pedagogical clarity, and 5) the
application of the typology should emphasise its status as a discourse of resistance to the
tradition of “either-or” approaches to translation decisions.

KEYWORDS

Translation solutions, translation pedagogy, translation procedures, theorisation of

translation, non-binarism.

1. Introduction

The naming of different ways of solving translation problems constitutes a

small European tradition that has spawned versions for languages beyond
Europe. A Romance-language strand can be traced from the work of the
Swiss linguist Charles Bally (1905) through Vinay and Darbelnet
(1972[1958]) and Vázquez-Ayora (1977). A Russian tradition might run from
somewhere around Fedorov (1927) through Retsker (1950, 2007[1974]),
Shveytser (1973), and Barkhudarov (1975), with Fedorov (2002[1953]) then
feeding into Loh (1958) at the beginning of a Chinese tradition. In Central
Europe we find typologies by Levý (2011[1963]) and Popovič (1968, 1975),
with Kade (1968) introducing concepts into German. Yet the most influential
categorisation undoubtedly remains Vinay and Darbelnet (1972[1958]), not
just in Romance languages but also in Japanese (for example in Hasegawa
2011) and elsewhere. The many more recent typologies include significant
attempts at systematisation by Delisle et al. (1999) and Schreiber (1998),
whose approach I partly draw on here. As a result of all this, we now find
numerous ways of identifying solution types, and we find them in virtually all
the major languages in which translators are trained. What we do not find,
though, is significant consensus on what these types are, how they should be
named, and whether they should concern more than one language pair.

41
The Journal of Specialised Translation Issue 30 – July 2018

In my study of that tradition (Pym 2016a), I opt for the general term
‘translation solutions’ (after Zabalbeascoa 2000) rather than ‘procedures,’
‘techniques,’ or ‘strategies.’ This is because the typologies I have just named
are all based on comparisons of texts, on the solutions adopted, rather than
on analysis of the cognitive processes involved. That minor point of
clarification does little to reduce confusion, however. I have sought to
discover (again in Pym 2016a) why there are so many different typologies,
and the answer seems to concern more than specific language pairs. The
plurality is partly due to different cultural conceptions of what translation is
and what relations between languages should be. I have also asked why so
much intellectual energy has been invested in these typologies, which have
constantly been modified without any significant input from cognitive
research. My premise here is that the typologies continue to be useful
pedagogically, specifically for making trainee translators aware of the range
of options available to them. From that perspective, the key way to improve
a typology is to see how well it opens options in the classroom.

In my 2016 book I try to synthesise the various pedagogical virtues of the

previous categorisations in a way that might be suitable for teaching
translation between many languages, including between European and Asian
languages. Table 1 presents the synthesis, incorporating a few subsequent
modifications. The synthesis tries to reduce contradictions, to explain things
in a clear and memorable way, and to invite students to think beyond their
initial or default solution types.

Here I report on how that typology has fared in Masters seminars given in
two very different contexts: Monterey (United States) and Stellenbosch
(South Africa). Both these contexts had been used for previous classes using
the more classical typologies and earlier versions of my model (on the
Monterey classes, see Pym and Torres-Simón 2015)1.

So my mission here is to identify the pedagogical strengths and weaknesses

of the resulting typology. If the typology is to fulfil its basic aim of making
novice translators extend their repertoire of available solutions, then we
would hope to find students engaging in discussion of that repertoire. In
qualitative terms, failure here is to be measured in terms of puzzled looks
and expressed misunderstandings, while success would be a reduction in the
puzzled looks and an increase in the students’ willingness to discuss and
extend the typology. Further, if the typology is to be applicable to many
languages, the seminars should be successful in both contexts.

I first present and explain the typology, then the results of the two seminars.

42
The Journal of Specialised Translation Issue 30 – July 2018

2. A typology explained

The typology to be tried out here has one default category: ‘cruise mode’
translating, as when an airplane is cruising at altitude; all goes well until
there is a ‘bump,’ attention is required, and something needs to be done. To
handle instances of ‘bump mode’ there are then eight main solution types
(the middle column in Table 1) that can be used for conscious problem
solving. The eight types are more or less in the tradition of Vinay and
Darbelnet (1972[1958]), going from simple to complex, from low-effort to
high-effort, from close-to-the-text to greater translatorial intervention. Here
I use initial capitals for the names of the types to indicate their status as
technical terms.

 Copying Words: ‘Transcription’ in the broadest sense, where items from

one language are brought across to another. This may be on the phonetic
level (e.g. Sp. fútbol), morphology (e.g. Sp. balompié, literally ‘ball-foot’)
or script (e.g. McDonald’s in all languages, alongside Rus. Макдоналдс or
Ar. ‫ماكدونالدز‬, for example).

 Copying Structure: Syntactic or compositional structures are brought

across from one language into another, as in Open mouth and lend voice
to tongue (from the television series Spartacus), where Latin syntactic
and metaphorical structures are used in English. This would be Calque in
Vinay and Darbelnet (1972[1958]).

 Perspective Change: An object is seen from a different point of view, as

in a hotel being Complet (‘full’) in French and has No Vacancies in English.
This is Modulation in Vinay and Darbelnet (1972[1958]). Here the
category is extended to include changes in footing (e.g. between the
formal and informal second person), non-obligatory switches between
passive and active structures, as when I see Mount Fuji is rendered as 富
士が見える [Fuji is visible] (Kanaya 2004), and the giving of a rival name to
the one object that does not thereby change position in time or place
(e.g. the use of Danzig for Gdańsk, indicating which side of the border, or
indeed of history, you are seeing it from).

 Density Change: There is a marked change in the amount of information

available in a given textual space. Translators can reduce textual density
by using solutions that spread information over a greater textual space,
using Explicitation, Generalisation, and Multiple Translation, as when the
one word Gemeinde is translated as the six words ‘Gemeinde, German
unit of local government’ (Newmark 1981: 31). Using the inverse
solutions can increase density.
43
The Journal of Specialised Translation Issue 30 – July 2018

 Resegmentation: The splitting or joining of sentences; re-paragraphing;

generally changing the order of text parts at sentence level or above.

 Compensation: A value is rendered with resources different from those

of the start text and in a textual position or linguistic level markedly
different from that in the start text 2 . A classic example is Fr. On se
tutoie… [‘We can use the intimate second-person pronoun…’] rendered as
My friends call me Bill… (Vinay and Darbelnet 1972[1958]: 190).
Translator’s notes and prefaces can be seen as forms of Compensation,
since they use a markedly different textual position. The cutting or
joining of sentences (Resegmentation) would not be Compensation
because mere rearrangement does not exploit different linguistic
resources.

 Cultural Correspondence: Different elements in different cultures are

presented as carrying out similar functions, as in the case of
corresponding idioms such as Give him an inch and he will take a mile
rendered in Chinese as 得寸进尺 [‘get inch, take foot’] (Loh 1958: 2.107) or
culture-specific items (currency units, measures, etc.). This broadly
covers what Vinay and Darbelnet termed équivalence and Adaptation
(1972[1958]). It applies to all instances where the corresponding
referents are held to be in different special or temporal locations, as
opposed to cases where the same referent is given different expressions
but remains in the one location. So the choice between Danzig and
Gdańsk is Perspective Change, since the referent city is presumed to
remain in the same place even while the politics vary, whereas cricket
rendered as baseball to express the common value ‘popular summer
sport’ is Cultural Correspondence, since the two referents are held to be
operative in different cultural locations.

 Text Tailoring: Semantic or performative material in the start text is

deleted, updated, or added to on the levels of both form and content.
This corresponds to what Schreiber (1998) calls Interlinguale Bearbeitung
(roughly ‘interlingual adaptation’).

Scarcely original in themselves, these categories are at least of a number

that can be adjusted to fit pedagogical needs. For maximum simplification,
there can be just three categories: Copying, Expression Change, Material
Change (in the left-hand column in Table 1). For more focused work, there
are open-ended lists of sub-types (as in the right-hand column, plus the
more detailed columns that could be added to the right of that).

44
The Journal of Specialised Translation Issue 30 – July 2018

Cruise mode (normal use of language skills, reference resources, parallel

texts, intuition — anything prior to bump mode — so no special solutions are
needed)
Copying Copying Words Copying sounds
Copying morphology
Copying script …
Copying Structure Copying prosodic features
Copying fixed phrases
Copying text structure …
Expression Perspective Change Changing sentence focus
Change Changing semantic focus
Changing voice
Renaming an object …
Density Change Generalisation / Specification
Explicitation / Implicitation
Multiple Translation…
Resegmentation Joining sentences
Cutting sentences
Re-paragraphing…
Compensation New level of expression
New place in text (notes,
paratexts) …
Cultural Corresponding idioms
Correspondence Corresponding units of
measurement, currency, etc.
Relocation of culture-specific
referents …
Material Change Text Tailoring Correction / censorship / updating
Omission of material
Addition of material …

Table 1. A typology of translation solution types for many languages (cf. Pym
2016a: 220)

The typology is based on the following principles:

 The names for the types are as common as possible: Given the difficulty
students have in remembering the differences between terms like
Transposition, Modulation, and Adaptation (see Pym and Torres-Simón
2015), I have used words that are as transparent as possible, even at the
expense of being inexact (Copying Words, for example, is very forced
when it has to cover various parts of words as well as ideograms —
45
The Journal of Specialised Translation Issue 30 – July 2018

‘copying language’ would be more accurate but the students tend to find it
too vague, applicable to whole paragraphs or pages, or suggestive of
dictation exercises).

 The solutions only concern translators’ transformations of text: The

typology does not deal with peripheral activities like finding information; it
does not cover skills that can be mastered by non-translators, such as
writing well; it does not purport to describe the thought processes used to
reach a particular solution.

 The typology concerns situations where a significant choice is to be made:

It does not deal with the application of obligatory rules or fixed
terminology.

 It concerns more than one language pair: The typology has not been
derived from a comparison of languages, although it draws on many that
have.

 It does not prescribe when particular solutions should be used: In

principle, all solution types can be used to solve all problems, with the
range limited in each particular case by the degree of effort required and
the relative risk of communicative failure (see Pym 2015, 2016a: 236ff.).

 It accepts conceptual overlaps: The typology recognises that the one

textual product can embody more than one solution type. For example,
Text Tailoring will normally bring about some kind of Density Change,
although Density Change in itself need not necessarily involve Text
Tailoring (since Explicitation, for example, theoretically does not add
actual content).

 Its purpose is purely pedagogical: The typology should be judged

successful when trainee translators and interpreters are able to grasp the
terms and use them to extend or refine their previous conception of the
translator’s task.

 It is open-ended: The degree of detail can be modified in accordance with

the pedagogical purpose at hand.

The typology used here is almost the same as the one published in Pym
(2016a, 220). The one major difference is the rank of Resegmentation, which
comes under Density Change in the book but was a separate high-order
category in the classes given (as in Table 1). This change is of some
importance because the splitting and joining of sentences is reported as
being more common between European and Asian languages than between
46
The Journal of Specialised Translation Issue 30 – July 2018

European languages (Sakamoto 2014, 184). In theory, the splitting of

sentences reduces syntactic load and can thus be seen as modifying text
density, which would place Resegmentation as a sub-category of Density
Change. Yet I suspected that my theorising on this point was too laboured
for the effective training of professional translators. With the theory,
Resegmentation is a legitimate part of Density Change. Without the theory, it
can be a stand-alone category, pedagogically reminding students that cutting
and joining sentences is something they might want to consider.

There were several more general features of the typology that I particularly
wanted to gain feedback on:

1. Can the typology add to the ways students solve problems? This
crucially concerns the top box in Table 1, where the solution types are
distinguished from run-of-the-mill ‘cruise mode’ translating. The
explicit purpose of the typology is to help solve conscious problems,
not to describe all the cognitive activities of translators. In effect, the
model says: When you are stuck, when you are cruising along and
then something goes ‘bump,’ here are some things you might want to
consider 3 . This rough and ready notion of ‘cruise mode’ will mean
different things for each level of training and indeed for each student,
and this variability in turn reduces or expands the kinds of examples
that make all the other categories meaningful. For example, if one
student is using phrase-level explicitation intuitively, as part of their
default translation practice, then they have no need to see this as a
specific solution type. Another student, however, whose initial
translation concept is highly literalist and shies away from phrase-level
explicitation, will find this solution type meaningful, at least as an
invitation to try out new ways of solving problems. The concept of
‘cruise’ mode is thus designed to enable the typology to address
students at very different levels. It introduces a pedagogical process-
based model into a typology that is otherwise based on product
analysis. I wanted to know if this would be understood by students.

2. Can students remember the main solution types? As noted, the table
has eight main types in the middle column. The number seems about
right for quick memorisation. If eight are not enough, the right-hand
column is available for the many other things that translators can do in
each particular situation, including combinations of the main solution
types. So would the students be able to remember the eight terms?
Would they be able to extend discussion into the more specific types in
the right-hand column?

47
The Journal of Specialised Translation Issue 30 – July 2018

3. Can students grasp the difference between Perspective and Density?

Since these were the only partly new terms, it was important to see if
anything was really made simpler by distinguishing between the
perspective from which you see an object (a hotel is full in French but
has No Vacancies in English, so it is a question of perspective) from the
degree of detail with which you see it (more detail should produce a
‘denser’ text). This difference could be elaborated with significant
theorisation, particularly concerning density (or granularity) as an
effect of how near or far one is from an object. But would such
explanations really make things clearer?

4. Will students accept that translation can involve radical changes to the
start text? The typology includes a major category (in the middle
column) called Text Tailoring, which involves adding material, omitting
material, and changing material (understood as things done or referred
to in the text, perhaps in the sense of the German Stoff). For some
translation concepts, material-changing solutions lie beyond the
translator’s mandate: one should only reproduce the material that is in
the start text. For those closer to Skopos theory, such changes are a
legitimate part of “translatorial action” (Holz-Mänttäri 1984). So how
would the students react? Would they tell me that translators have no
right to enter the territory of Text Tailoring?

3. Teaching the typology

I taught the basic typology in two two-hour classes at Masters level, one in
Monterey and the other in Stellenbosch. These sessions served to stimulate
discussion about the range of options available and the methodological
problems of classification. I am not suggesting that this or any other
typology should be reduced to just one two-hour session: the learning
difficulties revealed in this initial session should ideally be addressed in a
series of follow-up classes. To that extent, the classes were experiments in
which the students were co-researchers4.

The two groups were quite different with respect to level of prior studies,
language combinations, and language directionalities, and they would be
commenting on different texts. So there can be no suggestion that they
provided a solid basis for comparing specific group variables. But that is not
what was at stake. Since the typology was designed to be as robust as
possible, I was interested in pragmatically finding out which items would
work in both contexts and which would present specific difficulties. If specific
difficulties could be attributed to local factors, all well and good (there is no
reason not to compare how the groups fared). If not, then new ideas could
be necessary — this was never going to be a well-controlled empirical study.
48
The Journal of Specialised Translation Issue 30 – July 2018

3.1. The Monterey class

The first class was at the Middlebury Institute for International Studies at
Monterey (MIIS) on November 18 2015 as part of the translation practice
course (called ‘Translation Practicum’) in the second year of the Masters
program. The class had 18 students working with the following languages in
addition to English: Chinese (8), Japanese (3), Korean (2), French (4), and
German (1). None of the students reported having been taught a typology
like this previously, although the Chinese students had been exposed to the
similar metalanguage in Ye and Shi (2009).

Prior to the class the students were asked to bring at least two pages of a
translation they had completed previously. This could be a professional
translation or a translation done in another course.

The class then began with a warm-up exercise in which the students had to
suggest solutions for 18 problems (see Appendix 1), almost all of them
drawn from the previous literature on translation solutions. Their proposed
solutions were then compared and discussed in order to see if there was any
kind of common metalanguage already in general use. The phrase-level
problems were designed merely to provoke the need for a metalanguage;
they should not be misconstrued as a representation of all actual translation
problems.

Once it was clear that some basic terminological agreements would be

necessary for the discussion to proceed, the typology (Table 1) was
presented step by step, using the warm-up examples to explain each type,
with whatever discussions arose.

The second hour of the class was an activity to be completed by the students
working in pairs. The students had to go through two pages of their previous
translation, identifying and counting the types of solutions they had used.
They were then asked to give short answers to three questions:

- Is it easy to distinguish between the solution types? If not, why not?

- How do you think these categories could be improved?
- Which of these solutions tend not to be given by MT systems?

This last question followed on from previous classes that had introduced
postediting techniques. It will only be dealt with briefly here.

49
The Journal of Specialised Translation Issue 30 – July 2018

3.2. The Stellenbosch class

The class given in Stellenbosch took place on April 14 2016 and was more or
less the same except that the students were in the first year of their Masters.
There were 11 students, working with the following languages in addition to
English: Afrikaans (8), German (2), and isiXhosa (1). Since there were
introductory and organisational matters to attend to, the warm-up activity
was completed prior to the class and much of the assignment was completed
after the class. The warm-up activity did not include problem 18 (an
authentic example of censorable racism), on advice from the course
coordinator. None of the participants reported having studied similar
typologies previously. Some of the teaching process sought to correct
shortcomings detected in the Monterey class, especially with respect to the
explicit problem-solving status of the solutions.

4. Results

4.1. General questions

The reported difficulty of the solution types was similar for both the student
groups, as shown in Figure 1. The question here was, “Is it easy to
distinguish between the solution types? If not, why not?”. The responses
suggest the typology was generally quite difficult to apply, although many
students went beyond simple Yes/No answers, indicating that some types
were more difficult than others (in which cases their answers have been
classified as “Depends”). The numbers suggest that things might have been
a little easier at Stellenbosch (since the teaching had possibly been
improved), but the number of students is too low for anything to be read into
the result. One Stellenbosch student complained that “the lecture we had on
it was quite short and we hurried through the slides,” while the Monterey
students made no similar comment. It is important to note that the relative
difficulty here concerns identifying the types used in the student’s previous
translation, not in remembering a series of definitions and fabricated
examples. In same cases, the nature of the previous translation made things
relatively easy. For example, a Stellenbosch student commented, “I found it
relatively easy to distinguish between the solution types, mainly because the
text that I have chosen had to be domesticated for a South African reader.
Thus my text consists mostly of cultural correspondence.”

A repeated comment, in both the Monterey and Stellenbosch groups, was

that the categories overlapped. For instance, from Monterey: “Solutions such
as density changes, resegmentation, and compensation can all overlap in a
translation.”

50
The Journal of Specialised Translation Issue 30 – July 2018

60%

40%
Monterey
20%
Stellenbosch
0%
Easy
Depends
Difficult

Figure 1. Is it easy to distinguish between the solution types?

As to the reported difficulty of each solution type, the general pattern is

partly similar (Figure 2), albeit with some clear differences. The Stellenbosch
class reported more difficulties with the concepts of Perspective Change and
Density Change, which is of interest because Density Change was
hypothesised as being more important in their cultural context (in view of the
social need to reach readerships with lower education levels), whereas the
Monterey class reported marginally greater difficulty with Copying Words
(hypothesised to be a more active category for movements between
European and Asian languages).

35%
30%
25%
20%
15%
10%
5%
0%
Monterey
Stellenbosch

Figure 2. Solution types reported to be difficult (percentages of difficulties

mentioned)

51
The Journal of Specialised Translation Issue 30 – July 2018

The responses were similar enough in the two groups, as might be expected
from a simple awareness-raising question. The students thought that
machine translation handles the ‘copying’ solutions well enough, but they
generally did not think it applied the rest. That perception of machine
translation may be erroneous, of course, since neural systems can come up
with some very apt examples of Cultural Correspondence. In general,
though, students got the basic message that these are all things that human
translators can be particularly good at — and should thus fear little from
machines, if and when they know how to apply the fancier solution types.

4.2. Difficulties with specific solution types

The students were asked to say whether it was easy to distinguish between
the solution types and how the categories could be improved (see Appendix
2). Their written replies provide much richer information on the various
sources of difficulty. Here I look at each major solution type, attempting to
gauge what the causes of the complaints might be, whether the complaints
can be remedied easily, or whether they constitute grounds for major
modifications of the typology or teaching method.

4.2.1. Copying Words

A few comments from Monterey indicate a serious misunderstanding: some

students saw the typology as purporting to describe everything in a
translation, or everything that translators do, rather than just a set of ways
to solve problems that occur in ‘bump’ mode, when the normal flow of
translation activity does not provide a solution. For example, one Monterey
student complained that “‘copying words’ is a little nonsensical for closely
related languages like English and French, which have a large number of
cognates.” The complaint would be entirely justified if the typology had set
out to account for all the words in a translation. In this case, the instructor
need merely stress that the copying in question strictly concerns words that
are not in the standard repertoire of the target language. This is precisely
what I did in the Stellenbosch class, where the feedback included no
indications of this kind of misunderstanding. As mentioned, the teaching
hopefully improved slightly.

A similar complaint, which is repeated for several of the solution types and in
both groups, is that the types overlap. This would be problematic if there
were two names for exactly the same solution, but in most cases it turns out
that more than one solution is being used in the same sentence, which is
only to be expected. In some cases the apparent difficulty is that two of the
main solution types are being combined to give a particular solution, which is
also fine (the “Gemeinde” example from Newmark combines Copying Words
52
The Journal of Specialised Translation Issue 30 – July 2018

with Density Change, for example). Where the issue of overlap becomes
more problematic is in cases like the following, from a Chinese student at
Monterey:

Sometimes in one sentence, several types can be identified. For example, in

translation, we often re-segment the original long sentence into several short ones, so
basically resegmentation is used. But in those short sentences, we adopt the literal
translation, i.e. copying words. So at least two solution types can be identified.

The problem here is that Copying Words is misunderstood as literalism or

translating word-for-word — such references to literalism were very
problematic among the Chinese students reported on in Pym and Torres-
Simón (2015). Once again, greater emphasis has to be placed on the status
of the solution types as coming into play when standard language usages
and translation procedures fail to solve the problem. For this reason, “literal
translation” is not one of the categories here (whereas it is in Vinay and
Darbelnet 1972[1958]), although it might return in future versions (as a way
of translating irresolvable ambiguity, for example).

A much more astute complaint comes from a student of Japanese at

Monterey:

Japanese copies a lot of English words, and then chops them up so they aren’t so long
in Japanese (because every consonant is followed by a vowel in Japanese, consonant-
heavy English becomes many syllables longer in Japanese), so I wasn’t sure if that still
counted as copying words as they would be largely unrecognisable to native English-
speakers.

The point seems quite valid. The response in this case might be to add
‘hybrid word formation’ or somesuch to the list of possibilities in the right-
hand column. There is no need to exclude language-specific solution types.

4.2.2. Copying Structure

The use of the very broad term ‘structure’ was mentioned in many of the
general comments on the difficulty of the categories, but no specific
suggestions were made for improvement. The sense becomes clear enough if
one gives examples of things like syntactic calque and reproduction of rhyme
schemes. Hopefully the examples will compensate for the imprecision of the
name.

4.2.3. Perspective Change

As indicated, Perspective Change is an alternative name for Vinay and

Darbelnet’s musical metaphor “modulation” (1972[1958]). It seemed to work

53
The Journal of Specialised Translation Issue 30 – July 2018

quite well for the standard solutions: conversions of the ‘half-full half-empty’
kind, transformation of a positive into a negated negative, and switches
between active and passive voices — all of these on the semantic and/or
syntactic levels. The going becomes more difficult, however, when we try to
extend ‘voice’ beyond the active vs. passive opposition, minimally to include
subjectless verbs (particularly in Japanese) and changes in formal/informal
persons, but also potentially to see ‘voice’ at work in any feature changing
who speaks to whom in translations (cf. Sakai 1997; Alvstad and Rosa
2015). How far one wants to go down that road is a matter of context. In my
classes I have tried to stick to voice shifts that concern the status of first and
second persons and that clearly count as “seeing the same thing from a
different perspective.” The problem is that these days a whole lot of other
things can be included under the concept of ‘voice,’ which is probably why
Perspective Change scored highly for difficulty of application, especially in
Stellenbosch (Table 2).

This problem of expansive definition was picked up in several general

comments, particularly in the following observation from Stellenbosch:

If for example, a formal text is reworked for a newspaper, then some of the formal
expressions used will have to be adapted to suit the audience, which will be your
everyday newspaper readers. Even though the content would not be changed, the way
the author conveys the information will be changed. Will this then be a perspective
change (change of voice) or will it belong to a whole new category?

Good question! I think the actual change of text type is operating on a whole
different level (it is a macro-decision about how to write, made between
translator and client with respect to purpose); it is not a solution to a single
translation problem as such. Once you decide to change text type, though,
that decision can give rise to a series of smaller-scale translation solutions,
potentially including all eight of our main categories.

4.2.4. Density Change

The concept of Density Change was perhaps the major innovation with
respect to previous typologies. It was also one of the most problematic,
judging by the numbers in Table 2. There were several requests for further
explanation.

I suspect the confusion was due to a basic mistake in my teaching rather

than in the category as such. In the interests of making the types easy to
remember, I proposed that Perspective Change concerns the angle from
which you see an object, whereas Density Change concerns how close you
are to the object. The visual metaphor should make some sense: the closer
you are to an object, the more detail you see in your field of vision (there is
54
The Journal of Specialised Translation Issue 30 – July 2018

greater granularity); as you move further away, your field of vision includes
more general things. So the metaphor should help explain solutions like
Generalisation/Specification or Explicitation/Implicitation. At the same time,
ideally, it could encourage students to consider factors like how far the new
reader is from the object. Unfortunately, this turned out to be one of those
things that, the more you try to explain it, the more puzzled looks you get
from students. There were smart questions along the lines of, “If you are
close to something, then you know it well, so you leave a lot of information
implicit, and this gives less detail, surely?”. Hmm…

I would now suggest not mixing a visual metaphor (how close you are to an
object) with quantitative analysis (density is technically the cognitive load
required to process a stretch of language). Let us simply indicate to students
that there are several ways of stretching information over more textual
space, and that this is a way of making texts more accessible to new readers
or more economical to expert readers.

There were few other suggestions for improvement. A Monterey student

added, “I also think it’s important to emphasise that density can go both
ways, depending on the text and source and target languages.” Another
Monterey student brought up a further case of overlapping categories:

For Density Change, if expansion (explicitation) and contraction (implicitation) belong

to this category, how about using cultural-specific idioms to condense the original
words? Sometimes a long English sentence has a very succinct Chinese idiom
equivalence, and when the idiom is used, it’s both Density Change and Cultural
Correspondence.

Indeed, there can be two good reasons for adopting the one decision.

4.2.5. Resegmentation

There were no suggestions concerning Resegmentation, and it seemed

relatively easy to apply (Figure 2). The concept is something the Asian-
language students were already familiar with, and there were no doubts
about it in Stellenbosch.

This should justify separating Resegmentation from Density Change, since

the latter is already a handful without it.

4.2.6. Compensation

Although a standard item in most typologies, Compensation was reported as

being one of the more difficult categories to apply (Figure 2). The main
source of difficulty seems to be the requirement that the solution be on a
55
The Journal of Specialised Translation Issue 30 – July 2018

different linguistic level or in a different textual place. The meaning of ‘level’

probably required some working knowledge of linguistics, while the sense of
‘place’ has all the problems of any continuous variable: exactly how far away
do you have to be if you are in a ‘different’ place? The basic message
nevertheless seems clear enough: feel free to put the solution elsewhere.
Perhaps because of that very simple principle, there were no suggestions as
to how to improve the category.

4.2.7. Cultural Correspondence

The category of Cultural Correspondence brings together Vinay and

Darbelnet’s adaptation (cricket corresponds to baseball) and équivalence
(corresponding idioms), to which we might add all corresponding realia
(monetary systems, weights and measures, date conventions, etc.). Since
there are generally few such solutions on the cultural level and the
measurement realia are fairly straightforward, the application of the category
was reportedly quite easy (Figure 2).

There were just two suggested improvements, both of them indicative of

some degree of misunderstanding. A Stellenbosch student commented, “I
think cultural correspondence could be divided into more specific categories,
like domestication and globalisation.” This could indeed be done, and the
placing of ‘globalisation’ alongside or instead of the more traditional
‘foreignisation’ could even be stimulating: we could suggest that solutions
are referentially located in the start culture, in the target culture, or vaguely
‘elsewhere’ (see Pym 2010[1992]: 64-65). The initial purpose of the whole
typology, though, was to provide a mode of thought that could be an
alternative to the binarism of ‘either/or’ solutions. The binarism could be
applied, but there are reasons not to. It is equally possible to open up a
continuum of locations or an intellectual space in which solutions could be
tagged at any point (cf. Hervey and Higgins 1992: 33; Pym 2016a: 235).

The second suggested improvement was from Monterey: “Instances of name

adaptation in translation don’t seem to really have a place in the current list
of categories.” As mentioned above, this can be handled by distinguishing
between cases where the referent is held to remain in the same position in
place and time, when we are thereby concerned with Perspective Change
(the city does not move, after all), and cases where there are two different
cultural locations involved (cricket and baseball are held to be in different
cultures, as indeed would be metric vs. imperial measures, and so on).

This whole issue, though, depends very much on what is considered normal
or ‘cruise mode’ for a particular student or indeed translation culture. Going
into Spanish, the default rendition of London is Londres and Prince Charles
56
The Journal of Specialised Translation Issue 30 – July 2018

quite normally becomes el príncipe Carlos, and there is no particular solution

type required. At the same time, if a translator chooses to transgress the
norm and render Prince Charles as a Spanish Prince Charles, that special
solution is necessary marked and must be considered a case of Perspective
Change. Or again, consider Barcelona, which savvy English-speaking
travellers occasionally like to pronounce with a theta (barθeˈlona), somehow
positioning their knowledge of Spanish as being particularly intimate or local.
Of course, they get it politically wrong: the Catalan pronunciation is with an
/s/ (bəɾsəˈlonə) rather than a /θ/, which leaves the English-speakers taking a
side they probably know little about.

Something has to go bump before a solution type is called for.

4.2.8. Text Tailoring

My main worry about Text Tailoring was that some students would insist that
a translator has no right to add to or remove material. That did not happen,
remarkably. The reasons for the apparent acquiescence remain mysterious:
it may be due to students having misunderstood the radical nature of the
concept, or perhaps to their keenness to please the professor by finding all
solution types, or even to a historical tendency among younger translators to
assume a greater right to intervene.

Whatever the case, the main concerns here were once again with overlap. A
Monterey student commented, “When translating from English to Chinese,
omitting content may be to change the semantic focus. At the same time,
the density of the sentence is also changed. In this way, it is hard to say
which method I exactly adopted.” This sounds logical: the deletion of
material can reduce the density of the translation. Then again, two things are
being mixed here: Density Change is supposed to be when no actual content
is deleted: information is just spread over more or less textual space. On the
other hand, the kind of deletion involved in Text Tailoring is when something
is really cut out, both as content and as form . If the deletion is on all
levels, then it need not affect density: the remaining material is spread over
more or less the same textual space as it had in the start text. So I see no
need to concede to overlapping categories on that score. A Stellenbosch
student seemed to agree with this conclusion: “I categorised my solutions by
making use of these definitions: Explicitation is when you elaborate in the TT
on what has been said in the ST, and addition of material is when you add
something to the TT which was never mentioned in the ST.”

57
The Journal of Specialised Translation Issue 30 – July 2018

4.2.9. Cruise mode

That discussion effectively underscores what it means to have a variable

concept like ‘cruise mode,’ which necessarily involves quite different things
for different people, and indeed for different language pairs. Surely we
should go back to typologies that are specific to each language pair? What is
the real price being paid for a shot at universalism?

Such concerns are understandable but not insurmountable. Surely love is

different for each person and perhaps in each culture, but it is still something
worth naming and talking about? In each particular case, for each particular
potential lover and in each particular situation, you consider carefully how to
act. In the same way, every translation teacher is constantly monitoring the
solutions students use by default, and should ideally be pitching their lessons
at a level just one or two steps beyond that, in a “zone of proximal
development” (Vygotsky 1978: 84ff.). If French students are being too
literalist and need to be reminded that Transposition is legitimate and
normal, then you elevate that to a main item to be taught. And if Chinese
students are unsure of when they should use transliteration or phonetic
reconstruction of non-Chinese names, then that is a topic that should be
dealt with at length with them. Yet those differences do not mean the French
and Chinese students cannot share a common metalanguage for talking
about what they do. Just like love, the differences need not stop us talking
about it.

Doubts about the cruise concept ensued from an apparent desire to have
categories that have the same borders for all users. In the world of
translation, that is rarely going to happen. This is due to the way translation
performs its borders. Just as the boundaries of cruise mode are wherever
cognitive processes go bump, so the borders between languages and
between cultures are where translators intervene to mark the two sides of
their event. Or again, the border between Perspective Change and Cultural
Correspondence lies, in the case of names, in the way the translational act
itself positions the references in play, in one place or in several. Since the
basic borders are performed anew by each translation event, none can be
considered eternal.

5. Conclusions

The negative comments could make our typology look like a complete
disaster. There was nevertheless a comforting number of positive comments
along the lines of “The categories are well-organised,” “Like them as is,” “No
suggestions for improvement,” “Generally clear and applicable,” and “It is
rather easy to distinguish between the solution types.” Further, the students
58
The Journal of Specialised Translation Issue 30 – July 2018

tended to remember the names of the main solution types reasonably well.
There is pressing no need to wipe the slate clean and start again.

Of course, there are still some areas in which the typology needs
improvement, or at least careful attention when teaching:

1. It is crucial that the typology not be misconstrued as a description of

everything that translators do (see 4.2.1, 4.2.7, 4.2.9 above). When
translators are working in ‘cruise’ mode, there is no need for any
elaborate typology of solutions. When, however, a problem appears for
which no solution seems available, the typology proposes a set of solution
types that the translator might want to consider. Its utility is thus purely
pedagogical.

2. The lists that appear to the right of the central terms in Table 1 are open-
ended, so new solution types can be added in accordance with the focus
of a particular text or lesson (see 4.2.1, 4.2.7). Name Adaptation, for
example, is easily slotted in under Cultural Correspondence.

3. The concept of Density Change should be kept as simple as possible (see

4.2.4). Density merely concerns the amount of information in a given
textual space. The reduction of density, usually by using more textual
space, can make the translation more accessible.

4. For the same basic pedagogical simplicity, Resegmentation should be

presented as a category separate from Density Change (see 4.2.5).

5. Many of the categories can be combined to produce specific solutions, and

some can give two or more good reasons for the same solution (see
4.2.1, 4.2.2, 4.2.4, 4.2.8). In this sense, overlaps need not be seen as
shortcomings.

6. Much as binary thought underlies some of the decisions that are possible
when translating, the main virtue of the typology is that it presents an
alternative to binarism (4.2.7).

Even with these possible improvements and clarifications, the typology

should certainly not be considered definitive. It is no more than a suggested
teaching aid, open to adaptation to the students’ level and languages, and
the focus of each particular lesson. Perhaps the most unanticipated point in
the above discussion is that effective pedagogy can be hampered by
excessive theory (I gave up trying to insist that Resegmentation changes
density).

59
The Journal of Specialised Translation Issue 30 – July 2018

One final parting shot at binarism: Contemporary translation theory has very
little time for complex typologies of what translators do. Ever since Cicero
opposed “ut interpres” to “ut orator” as two modes of translation, positing
that one translate either like an oral mediator in a negotiation or like a public
orator, we have had a slow-grinding discussion of one side against the other,
often framed as better against worse. Our all-encompassing cultural theorists
love that kind of debate, and a hundred in-class discussions can be initiated
by pitting one side against the other, in the firm knowledge that neither side
will ever emerge wholly victorious. In such debates, students are learning
about translation, or about thought on translation, but not in a way that is in
close contact with actual translation practice. From that perspective, my
concern with translation solutions is definitely regressive: I am going back to
boring old linguistics; I am working close to actual texts; I am stirring up
issues on which virtually no empirical advances have been made.

Sooner or later, though, someone will want to learn how to translate. And a
widened repertoire of translation solutions, with more than just two major
terms, is one of the most valuable aids we can offer them.

References

 Alvstad, Cecilia, and Alexandra Assis Rosa (2015). Voice in retranslation. Special
issue of Target 27(1). Amsterdam/Philadelphia: John Benjamins.
 Bally, Charles (1905). Précis de stylistique: Esquisse d’une méthode fondée sur l’étude
du français moderne. Geneva: A. Eggiman.
 Barkhudarov, Leonid S. (1975). Язык и перевод (Вопросы общей и частной теории
перевода) [Language and translation: Questions in the general and specialised theory of
translation]. Moscow: International Relations.
 Bowker, Lynne (2016) “The Need for Speed! Experimenting with ‘Speed Training’ in the
Scientific/Technical Translation Classroom.” Meta. Translators’ Journal 61: 22-36.
 Delisle, Jean, Hannelore Lee-Jahnke, Monique Catherine Cormier and Jörn
Albrecht (1999). Terminologie de la traduction. Amsterdam/Philadelphia: John
Benjamins.
 Fedorov, Andrey V. (1927). “Проблема стихотворного перевода” [Problems in the
translation of poetry]. Poetika 2: 104-118.
 - (2002[1953]). Основы общей теории перевода (лингвистические проблемы) [The
general theory of translation (linguistic problems)]. 5th ed. Saint Petersburg: St.
Petersburg State University Faculty of Philology; Moscow: Filologiya tri.
 Hasegawa, Yoko (2011). The Routledge Course in Japanese Translation. London/New
York: Routledge.
 Hervey, Sándor and Ian Higgins (1992). Thinking Translation: A Course in Translation
Method: French to English. London/New York: Routledge.

60
The Journal of Specialised Translation Issue 30 – July 2018

 Holz-Mänttäri, Justa (1984). Translatorisches Handeln. Theorie und Methode. Helsinki:

Academiae Scientarum Fennicae.
 House, Juliane (1986). “Acquiring translational competence in interaction.” Juliane
House and Shoshana Blum-Kulka (eds) (1986). Interlingual and Intercultural
Communication. Tübingen: Narr, 179-191
 - (2000). “Consciousness and the strategic use of aids in translation.” Sonja Tirkkonen-
Condit and Riitta Jääskeläinen (eds) (2000). Tapping and Mapping the Processes of
Translation and Interpreting: Outlooks on Empirical Research. Amsterdam and
Philadelphia: John Benjamins, 149-162.
 Kade, Otto (1968). Zufall und Gesetzmäßigkeit in der Übersetzung, Leipzig: VEB Verlag
Enzyklopädie.
 Kumaravadivelu, Bala (1994). “The Postmethod condition: (E)merging strategies for
second/foreign language teaching.” TESOL Quarterly 28(1): 27-48.

 Kanaya, Takehiro (2004). 英語にも主語はなかった [English Used to Have No Grammatical

Subject]. Tokyo: Kodansha.
 Levý, Jiří (2011[1963]). The Art of Translation. Translated by Patrick Corness, ed. by
Zuzana Jettmarová. Amsterdam/Philadelphia: John Benjamins.

 Loh, Dian-Yang (陆殿扬) (1958). 英汉翻译的理论与技巧 / Translation: Its Principles and

Techniques. 2 vols. Beijing: Times Publishing.
 Mossop, Brian (1995). “Understanding poorly written source texts.” Terminology Update
28(2): 4-21. goo.gl/x8Ya0O (accessed November 2016).
 Newmark, Peter (1981). Approaches to Translation. Oxford: Pergamon Press.
 Popovič, Anton (1968). Preklad a výraz [Translation and expression]. Bratislava: VSAV.
 - (1975). Teória umeleckého prekladu: Aspekty textu a literárnej metakomunikácie
[Theory of artistic translation: aspects of text and literary metacommunication].
Bratislava: Tatran.
 Prabhu, N. S. (1990). “There is no best method – why?” TESOL Quarterly 24(2): 161-
176.
 Pym, Anthony (2010[1992]). Translation and text transfer. Tarragona: Intercultural
Studies Group.
 - (2009). “Using process studies in translator training. Self-discovery through lousy
experiments.” Susanne Göpferich, Fabio Alves, and Inger M. Mees (eds) (2009).
Methodology, Technology and Innovation in Translation Process Research. Copenhagen:
Samfundslitteratur, 135-156.
 - (2015). “Translating as risk management.” Journal of Pragmatics 85: 67-80.
 - (2016a). Translation solutions for many languages. Histories of a flawed dream.
London: Bloomsbury.
 - (2016b). “A spirited defense of a certain empiricism in Translation Studies (and in
anything else concerning the study of cultures)”. Translation Spaces 5(2): 289–313.
 Pym, Anthony and Torres-Simón Esther (2015). “The pedagogical value of translation
solution types.” Perspectives: Studies in Translatology 23(1): 89-106.

61
The Journal of Specialised Translation Issue 30 – July 2018

 Retsker, Yakob I. (1950). “О Закономерных соответствиях при переводе на родной

язык [On regular correspondences in translation into the vernacular].” Klavdiya A.
Ganshina and I. V. Karpov (eds) (1950). Вопросы теории и методики учебного
перевода [Questions of the theory and methodology of translation teaching]. Moscow:
Akademiya pedagogicheskikh nauk RSFSR, 156-183.
 - (2007[1974]). Теория перевода и переводческая практика [Theory of translation and
translation practice]. Moscow: R. Valent.
 Robinson, Douglas (1997). Becoming a Translator. An Accelerated Course. London/New
York: Routledge.
 Sakai, Naoki (1997). Translation and subjectivity. On “Japan” and cultural nationalism.
Minneapolis: University of Minnesota Press.
 Sakamoto, Akiko (2014). Translators theorising translation: A study of Japanese/English
translators’ accounts of dispute situations and its implications for translation pedagogy.
PhD thesis. University of Leicester.
 Schreiber, Michael (1998). “Übersetzungstypen und Übersetzungsverfahren.” Mary
Snell-Hornby, Hans G. Hönig, Paul Kussmaul and Peter A. Schmitt (eds) (1998) Handbuch
Translation. Tübingen: Stauffenburg, 151-154.
 Shveytser, A. D. (1973). Перевод и лингвистика [Translation and linguistics]. Moscow:
Voyennoye Izdatelstvo.
 Stern, Hans Heinrich (1983). Fundamental Concepts of Language Teaching. Oxford:
Oxford University Press.
 Vázquez-Ayora, Gerardo (1977). Introducción a la traductología: Curso básico de
traducción. Washington: Georgetown University Press.
 Vinay, Jean-Paul and Jean Darbelnet ([1972]1958). Stylistique comparée du français
et de l’anglais: méthode de traduction, Nouvelle édition revue et corrigée. Paris: Didier.
 Vygotsky, L. S. (1978). “Interaction between learning and development,” translated by
M. Lopez-Morillas. Michael Cole, Vera John-Steiner, Sylvia Scribner and Ellen Souberman
(eds) (1978). Mind in society: The development of higher psychological processes.
Cambridge, MA: Harvard University Press, 79-91.
 Wallace, Melissa (2014). “Team-based learning in introductory translation courses.”
Ying Cui and Wei Zhao (eds) (2014). Handbook of research on teaching methods in
language translation and interpretation. Hershey, PA: Information Science Reference, 27-
45.
 Ye Zinan and Shi Lynette Xiaojing (2009). Introduction to Chinese-English translation.
New York: Hippocrene Books.
 Zabalbeascoa, Patrick (2000). “From techniques of translation to types of solutions.”
Allison Beeby, Doris Ensinger and Marisa Presas (eds) (2000). Investigating translation.
Amsterdam/Philadelphia: John Benjamins, 117-127.

62
The Journal of Specialised Translation Issue 30 – July 2018

APPENDICES

1. Warm-up activity

Please translate these into your favourite language other than English,
except the ones that are not in English – they should go into English, if you
can.
1. Segway
2. Zeitgeist
3. Hoverboard
4. Skyscraper
5. Translation Studies
6. 富士が見える
7. Danger de mort
8. Lebensgefahr
9. No vacancies
10. Gemeinde
11. Real IRA
12. The final whistle (in a soccer game)
13. [Here just indicate the problem and how you would solve it.] The Spanish
Government retains the sole property rights to this document without
prejudice to rights of the holder and therefore recommends that the holder
take the utmost care regarding the custody and use of the passport and
requests that whatsoever authority or other person return the passport to
the Spanish authorities should it be lost or used in an unjustified way.
14. “You can call me Bill.”
15. Pearls before swine.
16. Before you could say Jack Robinson
17. There is a 327-meter freeway between the two cities.
18. The white man has led human progress for two thousand years.

2. Class assignment

Working in pairs if possible, please take two pages of a translation you have
done previously and try to identify the main solution types. Say how many
instances you found of the following:

Copying Words
Copying Structure
Perspective Change
Density Change
Resegmentation
Compensation
Cultural Correspondence
63
The Journal of Specialised Translation Issue 30 – July 2018

Text Tailoring

Now briefly answer the following:

- Is it easy to distinguish between the solution types? If not, why not?
- How do you think these categories could be improved?

Biography

Anthony Pym was Visiting Professor at the Middlebury Institute of

International Studies at Monterey from 2008 to 2016 and has been Professor
Extra-ordinary at Stellenbosch University since 2011.
E-mail: [email protected] / [email protected]

Notes
1
A further outing was in an Academia session from September 13 to October 13, 2016
(https://www.academia.edu/s/f91e5c89bb/a-typology-of-translation-solution-types). From
that session I draw on feedback from Alexander Zaytsev, Sergio Portelli, Felix Larsson,
Dorota Guttfeld, Lily Robert-Foley, Douglas Robinson, Yves Gambier, and Nader Albkower, to
all of whom my sincere gratitude.
2
I prefer ‘start text’ to ‘source text’ because these days translators work not just from a
single text but also from glossaries, translation memories, and machine translation output;
any one of those resources could provide the ‘source’ for a solution. The role of the initial
text is thus relativised.
3
The notion of a ‘bump’ comes from Mossop: “In normal mode, people who have mastered
some skill simply ‘see’, instantly, how to proceed. In bump mode, however, principles have
to be applied” (1995: 4). The idea of a default mode is elaborated at some length in
Robinson (1997).
4
The use of research methods in the hands-on translation class is by no means new (see,
for example, House 1986, 2000, Pym 2009, Wallace 2014, Bowker 2016) and should not be
confused with empirical pre-testing prior to public release, as if solution typologies were like
a new drug or airplane. As with many complex ideological constructs (lists of translation
competences would be another example), we should recognise that the distinctions between
categories will always be to some extent arbitrary and/or contextually contingent. In these
cases there is no absolute truth to be found in empirical testing, and we should renounce
that white-coat objective fantasy. Since the function of such typologies is to facilitate
discussion, awareness, and learning, their prime testing is their teaching, and will always be:
64
The Journal of Specialised Translation Issue 30 – July 2018

my reports here are meant not to provide transferable truths, but to suggest ways in which
further classes, further practical experiments, can be carried out, and the typologies can be
modified in accordance with results – all through a series of admittedly “lousy experiments”
(Pym 2009). This need not involve, however, any relativist position where all typologies are
somehow considered equally valid and no empirical testing is sought, as might be one
reading of a “post-method condition” in language teaching (see Stern 1983, Prabhu 1990,
Kumaravadivelu 1994). On the contrary, all testing is always of interest, since our classes
can always be improved and ideas and innovations are always welcome. The one condition
for this very applied kind of empiricism is that we start from an initial veil of ignorance (Pym
2016b): we do not initially know which typology works best, but we are going to try to find
out.

65